Service load balancing in fog-based 5G radio access networks

Fog-based radio access networks (F-RAN) are poised to play a significant role in future 5^th generation (5G) cellular networks by harnessing the distributed resources of collaborative edge devices to deliver localized RAN services to the end-users. Through ingestion and processing of end-user tasks close to their sources, F-RAN has potential to meet the stringent latency and bandwidth requirements of 5G services and applications. Due to the multitude of edge devices with different resource capacities that can be selected by the fog access point (FAP) as service nodes to process an end-user task, an important issue is deciding which service node(s) should be assigned to process what tasks from endusers in the F-RAN. The tasks here refer to user processing tasks that would have normally performed by the FAP or remotely by the cloud. In this paper, we introduce the concept of virtual FAPs (v-FAPs) formed by a number of local devices such as WiFi access points, femtocell basestations and the more resource-rich end-user devices under the coverage and management of the FAP. This paper focuses on addressing the task assignment problem by proposing a service load balancing algorithm for the v-FAPs. We model the user tasks as a task graph and the service nodes as an edgeless service graph. We then formulate an optimization problem to find the optimal assignment of tasks with the objective of balancing loads at the service nodes.